Dear Dr. Scalo, We appreciate the prompt and thoughtful attention our paper has received from both you and the referee. We are pleased to resubmit our manuscript ApJL #21952 in modified form with this response. We have attempted to address the concerns of the referee and in our opinion the manscript is significantly improved. In revising the manuscript, we have also modified our analysis in ways that we believe improve the consistency of the manscript, are conservative given our current understanding of the data, and do not change the basic results of the paper. We describe those changes in detail below. Regarding your concern about the ages adopted for the stars in our paper, we have made the following changes. We have made explicit reference to the calibration of the calcium H&K emission line indices that we have adopted for the older stars in our paper (section 2, paragraph 1, sentance 7), the highly cited thesis of R.A. Donahue (1993). We have also explicitly remarked that errors in these age estimates are factors of x 0.5-2.0 (section 2, paragraph 1, sentance 8), and we comment on the effect errors in age have on our results (section 3, paragraph 2, last three sentances). Regarding the more general issue concerning the ages of stars studied as part of our project, revised stellar age determinations for the entire FEPS sample will be discussed in detail by Hillenbrand et al. (2008, in preparation) as mentioned in the present paper. In brief, a variety of age indicators such as coronal activity, chromospheric activity, stellar rotation, lithium abundance, and the Hertzsprung-Russell diagram are calibrated to open clusters and used to assess stellar ages between 3 Myr and 3 Gyr. FEPS has invested substantial effort in re-deriving calibrations and determining the effective range of stellar ages over which each may be applied. For example, as you note, the HR diagram is a useful age indicator only for stars which are securely pre-main sequence or post-main sequence. Conversely, in the younger (<30 Myr) pre-main sequence phases, the coronal and chromospheric activity indicators become saturated and rotation is still tied to the previous epoch of inner protoplanetary disk evolution. Although there is an extensive previous literature on all of these topics, for consistency and uniformity in treatment, the FEPS project has revisited many aspects of the calibration data, the calibration cluster ages, and the second parameter effects such as stellar mass -- all before conducting the age analysis on the FEPS sample itself. Then for each and every star in the FEPS sample we have considered all the available age indicators and developed a procedure for estimating the plausible age range and the likely age of each target. Although we should have introduced the results of this exercise into the refereed literature before the onslaught of Spitzer results, we have not done so. We therefore request permission to proceed with the current manuscript in advance of publication of the Hillenbrand et al. (2008) paper to be submitted later this year. Concerning your specific comments, we note the potential utility of the Pont and Eyer (2004) contribution on isochronal ages. However, we use this approach only for a very small fraction of our sample as few of the stars in our sample are older than 3 Gyr. We also note that it will be difficult to compare our results concerning the calcium H&K emission line indices to the calibration of Pace and Pasquini since they do not convert there results to the R'HK index used by the majority of the workers in the field. We tried to exclude stars with no companions in our sample and to our knowledge none of our targets has a white dwarf companion. We appreciate your taking the time to review our PASP paper to which the reader is referred in our current contribution. It was an oversight not to include a more appropriate reference concerning our baseline calcium H&K emission line calibration (Donahue et al. 1993) in that paper. Finally, it was not our intent to state definitively that there are no stars older than 2.5 Gyr in a volume-limited sample of stars surrounding the Sun, but merely to indicate that their density is decreased as you indicate. I can see how the statement in the PASP paper could be misinterpreted. As age estimates continue to improve and representative samples grow larger and more refined, it will be important to continue to revisit the age distribution of stars in the galactic disk with an eye toward refining the star formation rate in the galaxy (not to mention the field star IMF!). We value your input which has provided much food for thought. As instructed, we have downloaded a version of the manuscript from the electronic site and editted that version to insure that we have preserved LaTeX corrections made to our paper. We have also included email addresses for all authors as requested. We discuss our response to each point raised by the referee below. Sincerely, Michael R. Meyer *****************Response to the Referee************************************* We thank the referee for a thoughtful and prompt report. She/he raised a number of important issues and we believe the manuscript is considerably stronger for having addressed these points. Our responses to each point are given below. In addition, we bring to the referee's attention a few additional changes to the manuscript. To start with, two points of confusion in our data led to a complete re-analysis which is reflected in the current draft. The first was a clerical error in the parent source list for this analysis which led to removal of two sources from the unbiased sample (HD 72905 and HD 38207) and the addition of two FEPS targets to the list (HD 105 and Sco PMS 214). At issue was whether or not the sources were added to the FEPS project list on the basis of previously suspected IR excess emission or not. Because HD 38207 was identified as a target with 24 micron excess emission, its removal changed the quantitative results reported in our paper in a minor way. The second was more substantive. In the last paragraph of section 2 of the submitted paper, we discuss possible integration time dependent changes in flux calibration and (eroneously) stated that we did not apply those offsets. In fact, we did apply those offsets (ranging from 1-3 %) to the 24 micron data we analzed. Because those offsets are larger than determined by the instrument team (Englebracht et al. 2007, in press), we believe it is best to defer analysis including those offsets to a future paper which explains in greater detail how those offsets are derived. As a result, we have completely redone our analysis using data calibrated in the standard way. The noise in our data is increased very slightly due to this change, but this has only a minor affect on the outcome of our study. The text, table, and figures all reflect this revised analysis. We believe these results improve the consistency of our paper, and are conservative given our current understanding of the data. We apologize for these errors and have worked hard in the past few weeks to correct them. Finally, in response to comments from the editor, we have expanded our discussion of the stellar ages used in our analysis including: a) explicit reference to the calcium H&K emission line calibration used in deriving ages for our stars (sect 2, par 1); b) comment on the errors in age (sect 2, par 1); and c) the effect of these errors on our results (sect 3, last par). And now for our responses to the referee's helpful comments. ###################################################################### REFEREE: The second paragraph of the introduction would be clearer if it stated explicitly the relevance of tracing planetesimal belts to the preceding discussion on terrestrial planet formation, for example by continuing "... (Meyer et al. 2007) and can be used to provide constraints on planet formation processes around other stars." RESPONSE: We adopt a slightly modified version of this suggestion in the current manuscript. REFEREE: At the end of the first paragraph of section 2, it is stated that Av=0 was assumed for stars within 75pc, but not what was assumed for more distant stars. RESPONSE: A_v was left as a free parameter in the fits for more distant stars. We have altered the text to make this more clear. REFEREE: The analysis mentions that the expected 24/8 micron photospheric ratio is 0.1, yet the mean observed ratio for stars without excess is 0.117. This discrepancy needs to be discussed. The third paragraph of section 3 mentions a systematic 3% offset in the calibration of 24 micron fluxes which is not applied and presumably contributes to the discrepancy. However, the same paragraph mentions a <4% offset in the 24/8 ratio. It is not clear if this figure of <4% takes into account the systematic 3% offset for 24 micron fluxes and the systematic 2% offset for 8 micron fluxes. Even if this is an additional systematic offset it seems that a 17% offset from the photosphere models is a significant discrepancy. Also, it is not clear whether these systematic offsets apply uniformly across the sample, or whether there is a dependence on absolute flux level (or spectral type). If there was this would introduce a systematic error in the fraction of stars with detectable excess as a function of age, since brighter stars are older. It should be possible to assess this by looking at the mean 24/8 ratio of stars without excess as a function of 8 micron flux, but this is not obvious from figure 1 which would be clearer with a logarithmic x axis so that the numerous faint sources can be separated. The wording in the second paragraph of section 2 "minimum uncertainties of <1%" is also confusing. RESPONSE: The confusion concerning the expected ratio of 24/8 micron flux based on the models was generated by lack of precision in the submitted text. The original text referred to model predictions of "approximately 0.1". We now include explicitly the average model prediction of this ratio MIPS24/IRAC8 = 0.116, which is very close to the average flux ratio of the sources lacking excess after sigma-clipping. This level of agreement is somewhat surprising given the many possible reasons they could be offset including: a) objects with small (undetectable) excesses remaining in the sample; b) model uncertainties, particularly for stars with low gravity; and c) residual calibration offsets. Regarding possible systematic offsets in the color ratios as a function of source brightness, we now explicitly address this in the text. We searched for a flux dependent offset in the MIPS24/IRAC8 micron flux ratio, and found a small offset between sources brighter than and fainter than 128 mJy. This could be due to small integration time dependent offsets in calibration, which will be addressed in a future contribution. This discussion is reported in the text, towards the end of paragraph 1, section 3. We also comment in a footnote concerning how this offset could affect our results. We thank the referee for helping us improve the clarity and precision of this discussion. REFEREE: The authors rightly compare their 24/8 ratios to an expected value of 0.117 for photospheric emission, even though the photospheric models predict a ratio of 0.1. To avoid mis-interpretation of the observed flux ratio, it would be useful if table 1 also quoted the derived (F24-F24*)/F24* ratio = (f24/f8 - 0.117)/0.117. It would also be useful to quote the S/N of the detection. However, when doing this it is concerning that for 8 of the 31 sources in table 1 quoted as having an excess, that excess is only significant at the 2-3sigma level. This arises because the population without excess has a distribution with a sigma of 0.004 which is lower than the sigma of any of the individual measurements (lowest sigma is 0.006 in table 1). The origin of this discrepancy needs to be resolved if we are to believe the detection of excess for sources which formally only have a significance of 2-3sigma, which otherwise should be removed from the analysis. RESPONSE: We have adopted the suggestion of the referee to quote the fractional 24 micron excess as defined in the paper and in the comment above, as well as report the uncertainty in this quantity. In the previous submission where we were quoting the flux ratio alone, we adopted an overly conservative definition of error in each flux measurement that included a calibration uncertainty for both the IRAC8 and MIPS24 micron data leading to the quoted uncertainties of order 0.06 in the ratio. As the referee points out, the dispersion in the mean color ratio for sources lacking detectable excess (now 0.005, slightly larger based on the new data analysis) characterizes the major uncertainty in the reported fractional excesses (0.005 is 4.3 % of 0.117) and the internal random uncertainties in each measurement contribute a small fraction of the total quoted errors. REFEREE: While the paper mentions that a K-S test shows that the age distribution of the 22 weak excess sources is not inconsistent with that of the 31 stronger excess sources, it does not say how those age distributions compare with that of all stars in the sample. It would also be useful to know the number of stars in the different age bins, so that the fractions on Fig. 2 can be readily computed. RESPONSE: Based on the outcome of the new analysis, we have removed the discussion concerning the "22 weak excess sources" and updated the text accordingly. We have adopted the referee's suggestion to quote explicitly the number of stars in each bin, as well as the fraction with excess, explicitly in the text. REFEREE: The paper says that the results for open clusters and field stars are different but formally consistent. It would be useful if the differences could be quantified so that the reader can judge the signficance of the similarity (and 2x difference) in excess fraction quoted. RESPONSE: We thank the referee for this helpful suggestion and have adopted it in the revised text. REFEREE: The paper correctly identifies the key question of whether 24 micron excesses last a long time, or are short-lived. It goes on to suggest that the total fraction of sources that exhibit 24 micron excesses at some point during their life is 60%, since it may be possible to sum the fractions of stars with excesses over the range 3-300Myr. However, it is not clear what assumptions are being made here or what they are proposing as their interpretation of the results (or what the alternatives are). For example, it seems that one key assumption is that the transition from orderly to runaway growth is associated with a large, detectable, 24 micron excess, and that an excess is not seen at any other time. While this is mentioned a couple of paragraphs later, "During or shortly after...", it is key to the assertion of the paper and so it is important that the reader has this picture of the 24 micron evolution during terrestrial planet formation in mind from an early stage in the discussion. To justify summing over the 3-300Myr range it must also be assumed that the range of initial disk masses is x100, and that the most massive disks have transitions at 3Myr. These assumptions are not justified, but the factor of 100 in disk mass could, for example, by reference to observed proto-planetary disk mass distributions such as those of Andrews & Williams 2005. I note also that the figure of 60% calculated with the above assumptions is a lower limit, since the 10% of 30-100Myr stars that are seen to have 24 micron excesses would only include *all* stars that have a transition from orderly to runaway growth in the period 30-100Myr if that transition (and so the 24 micron excess) lasted for the whole 30-100Myr period. RESPONSE: We appreciate the referee's point and have tried to improve the logic with which we present one interpretation of these results: that the fractions in figure 2 can be summed for a total fraction of > 60 %. We have moved the sentance beginning "During or shortly after..." into the previous paragraph as suggested. We have also added a crude calculation of the expected duration of this phase of 24 micron emission, the result of which is < x3, the width of our age bins. We have also added a reference to Andrews and Williams (2005) as suggested to justify more fully the assertion that one might expect a range of x100 in the transition from orderly to chaotic growth. REFEREE: That paragraph also quotes a lack of decrease of "average detectable excess". No information is given in the paper to back up a lack of decrease in the average detectable excess (i.e., the mean (F24-F24*)/F24*), unless the reader makes their own plots from Table 1. Perhaps the authors mean the "fraction of stars with detectable excess" which can be inferred by eye from Fig. 2 to be consistent as not having any decrease in the 3-300Myr range. Whichever quantity it is that the authors are claiming not to be decreasing, this should be explicitly mentioned and the lack of decrease quantified. However, it is not clear to me that a constant fraction with detectable excess (or a constant average detectable excess) is consistent with the suggested scenario. I am inferring that the scenario is that stars form disks with a x100 range of disk masses which results in a range of times to get to the runaway growth transition, and that these disks are not bright at 24 micron except during that transition. In this scenario the fraction of stars with excess would depend on the distribution of initial disk masses. To be consistent with the observed lack of decrease of either of the observed quantities would further require that the 24 micron excesses that are produced during the transition are always of a similar (detectable) magnitude. This may be the case, but is not intuitively so, since a disk which transitions at 100-300Myr in this scenario is of intrinsically low mass (and so, perhaps, is also of low dust luminosity). RESPONSE: We have investigated whether or not the mean detectable excess, now reported in Table 1, evolves with time. There is no evidence that the mean excess detected around stars 3-30 Myr is different that the mean for stars 30-300 Myr: = 0.359 with sigma = 0.199 = 0.3458 with sigma = 0.116 However, we do not have much confidence in this quantitative comparison as it is based on samples which are dominated by upper limits. We report this result, along with a description of what is expected under our proposed scenario as pointed out by the referee, in paragraph 2 of section 4. We thank the referee for pointing out the inconsistency in the original logic of the paper. REFEREE: The comparison with the statistics of A stars will be an important and valuable part of this paper, since there are similarities in timescale and differences in magnitude of excess, yet this discussion is lacking at the moment. The 24 micron statistics for A stars can also be explained without invoking a terrestrial planet formation model, but as the steady state evolution of planetesimal belts (Wyatt et al. 2007, ApJ, 663, 1103). Perhaps this comparison was made in the missing sentences "excesses to However,"? RESPONSE: We appologize for the latex error that led to the missing phrase at the end of the sentence to which the referee refers. We have added a sentence comparing the fractional excesses between A stars and G stars as the referee suggests (section 4, par 3, last sentence). REFEREE: The paper finishes with a statement that planet searches will provide a critical test of their assertion. However, the paper does not extrapolate from the discussion of 24 micron excesses to say anything about the planets present in these systems. The 24 micron fluxes in the scenario say that planet formation has progressed to orderly growth, so perhaps the authors are extrapolating to make a prediction that all stars have terrestrial planets of size sufficient to be detected with the quoted planet searches. Whatever the prediction is that will test their interpretation of the evolution of 24 micron excesses, it is not clear. RESPONSE: We thank the referee for the suggestion to sharpen this point. We have added a remark to the end of par 2 in section 4 to the effect that under the proposed scenario, planets that form late from lower mass disks, will likely be smaller. We have also rewritten the last paragraph of section 4 and hope that our general point, that these observations are consistent with most sun-like stars forming terrestrial planets, is now more clear.